• No results found

Extending the privacy calculus model: The effect of privacy concerns on the continuance intention of fitness trackers

N/A
N/A
Protected

Academic year: 2021

Share "Extending the privacy calculus model: The effect of privacy concerns on the continuance intention of fitness trackers"

Copied!
43
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

Extending the privacy calculus model: The

effect of privacy concerns on the

continuance intention of fitness trackers

By

Mark Roke

University of Groningen Faculty of Economics and Business

MSc Business Administration: Strategic Innovation Management & Business Administration: Health

July 2020

Supervisor: Dr. E. Smailhodzic Co-assessor: Prof. Dr. Ir. K. van Ittersum

(2)

Abstract

Growing capabilities of technology have created many opportunities for society to use health information technology to gain insights in physical well-being. However, the rising popularity of these devices and the growing capabilities to collect personal data is not without consequences for our privacy. As personal health information is shared with commercial vendors of these devices, it is unsure who can access this information and where it will be used for. Therefore, this thesis investigates the effects of the perceived privacy risks and perceived benefits on the continuance intention among users of fitness trackers. Knowledge about this extends the privacy calculus model, as research is until now often limited to the adoption of fitness trackers. Findings have shown that users are aware of potential privacy risks, however, the perceived benefits are found to be more important to determine the continued use intention.

Keywords: health information technology, privacy calculus, perceived privacy risk, perceived

(3)

Table of contents

Introduction

4

Theoretical background

6

The adoption of health information technology………. 6

Similarities and differences between adoption and continuance of use…………...7

The privacy calculus theory………. 8

Continuance of use intention……….. 12

Hypotheses development……….14

The moderating effect of trust………. 15

Methods

19 Research design………... 19 Data collection………. 20

Results

22 Descriptive statistics……….……… 22 Hypotheses testing………... 25

Discussion

27 Theoretical implications……….. 27 Managerial implications……….. 29

Limitations and further research………. 30

Conclusion

31

References

32

(4)

INTRODUCTION

The impact of technology has been growing tremendously since the 1990s. The developments in information and communication technology (ICT) have created opportunities to communicate, share and collect data for any purpose at any moment (Oh et al., 2005). The field of healthcare technology is no exception, as health information technology (HIT) has an increasing fundamental role in the delivery of care. HIT is often added to traditional methods of treatments and has in some cases entirely replaced the work of healthcare professionals (Kellermann & Jones, 2013). Although it began as a tool for healthcare providers to store patient data, diagnosing or decision-making, consumers now benefit from these innovations due to the opportunities to self-manage their health. Although the industry of HIT may sound diverse, it is especially one market within this industry which has generated increasing awareness and has been growing rapidly over the last years: the market for wearable fitness trackers (Wolf, Polonetsky & Finch, 2015; Patterson, 2013). The rapid developments in technology led to the quick diffusion (Fereidooni, Frassetto & Miettinen, 2017). In this research, a fitness tracker is described as an electronic accessory with sensor features, used for the collection of health data on an individual level (Wolff, 2013). However, such innovations are not without consequences for our privacy, as it is difficult to track who is able to access the data and where it is used for (Dinev & Hart, 2006). The opportunities for personalised data collection are still growing, which results in the privacy of users being increasingly vulnerable (Li, Wu, Gao & Shi, 2016). Therefore, the purpose of this study is to create better understanding of the effects of perceived privacy risks of fitness trackers on user’s intention to continue the use of their tracker.

(5)

(Shen, 2019, p. 92). Research has shown that increased possibilities for data collection, leading to more personalised use of HIT, are positively related to perceived benefits for individuals. However, at the same time it increases the perceived privacy risks as a consequence of more personal data being collected (Sun, Wang, Shen & Zhang, 2015; Wang, Duong & Chen, 2016). It is unknown if privacy will indeed be violated, as there is a possibility that despite the perceived risk, no privacy harm has to be caused to the user. According to Li, Gupta, Zhang & Sarathy, (2014), this depends on the estimation of the user about the trustworthiness of the vendor of the tracker. Related to the risks, trust gives the user a feeling that the particular vendor would not disclose any personal information with those who are not authorised or have bad intentions. So, the feeling of trust reduces the negative effect of perceived privacy risks on the continuance intention (Akter, Ray & D’Ambra, 2013). It is the extent to which the user believes that the vendor will stick to their words. Related to the benefits, trust has the consequence that the actual provided information by the fitness tracker is reliable and can be perceived as accurate and correct and free from mistakes (Li, Gupta, Zhang & Sarathy, 2014). In other words, it is the extent to which the provided information can be trusted by the user. Trust is defined as a set of three beliefs (competence, reliability & safety) that reflect confidence that personal information submitted to ICT devices is not used opportunistically (McKnight et al., 2002). Whereas perceived risks and benefits describe a general perception of fitness trackers, trust is related to the entity the information is shared with.

Up to now, research on the effects of privacy risks has focused on the adoption of fitness trackers, but previous studies have not incorporated continuance of use while this has crucial implications for the fitness tracker market (Li et al., 2016). Currently, the influence of privacy risks on post-adopters remains unknown. Despite this lack of research, it is argued that privacy concerns are not restricted to individuals in the adoption phase, but also during post-adoption (Limayem, Cheung & Chan, 2003). This research will extent the privacy calculus model by testing the effect of perceived privacy risks on the continuance intention of fitness trackers. Outcomes are valuable because effectiveness of HIT depends on the continued use behaviour of users, as fitness trackers become effective for health improvement if they are used over a longer period of time (Jasperson, Carter & Zmud, 2005; Attig, Karp & Franke, 2018). Vendors feel the need to keep customers engaged as profitability comes from long-term users (Bhattacherjee, 2001; Son & Han, 2011). The specific objective of this thesis is to give answer to the following research questions: 1. To what extent is the continuance of use of fitness trackers influenced

(6)

problem of insecurity of future prospects for the market of fitness trackers because the consequences of privacy risks become clear (Shin & Wang, 2017). This insecurity is the consequence of a lack of knowledge about the effect of privacy risks on the willingness to continue the use of the technology. Outcomes of this research contribute to the knowledge about post-adoption behaviour of fitness tracker users but also shows whether more effort should be undertaken to secure the privacy of users. Now, large investments are made for the development of this technology but the problem that is related to insecurity of future prospects remains. Outcomes of this study will contribute to the understanding of the viability of the fitness tracker market because retention of users is vital to this (Karahanna, 1999; Bhattacherjee, 2001). The privacy issue is becoming increasingly important and, in recent years, attempts were undertaken to make individuals aware of the risks of sharing personal information (Ball, Ramim & Levy, 2015). This thesis will contribute to the theory about the growing issue of privacy risks, by extending the current literature about the effects of privacy concerns on the continued use of technology. The outcomes of this study are the start of more research on continued use of health technology, to create a better understanding of the potential of self-managed health devices.

This paper will be organised as follows: first a theory section will discuss the theoretical background and the development of hypotheses. Then, the methods section will discuss the research design and procedures for data collection, followed by the results of the research and lastly a discussion and conclusion.

THEORETICAL BACKGROUND

The adoption of health information technology

The body of literature regarding the effects of privacy concerns on the adoption of HIT is extensive. Literature has shown that individuals who consider adoption are aware of potential risks of privacy violation, despite they have not made use of the technology yet (Patel, Beckjord, Moser, Hughes & Hesse, 2015).

(7)

privacy risk and perceived benefits (Xu, Teo, Tan & Agarwal, 2009; Li et al., 2016). Potential breaches of security systems or bugs leading to unauthorised access to the information will lead to losses of privacy. Higher concerns about this leads to increased levels of perceived privacy risk (Dixon, 2005). Adopters are willing to accept this privacy risk of their privacy because they are incentivised by the perceived benefits of adoption. So, in case of an adoption decision, disclosing more information demands higher perceived returns, as more information disclosure leads to higher perceived risks. The perceived privacy risk is only accepted when the expected outcomes of benefits are higher (Smith, Dinev & Xu, 2011).

Literature on adoption shows that the strong feelings of privacy risk can be counterbalanced by the benefits of use, if the individuals’ beliefs about benefits surpass their perceived risks (Kim, Ferrin & Rao, 2008). Comparing adoption with the alternative, no adoption, will ultimately lead to a positive or negative attitude about the adoption of HIT. Attitude as a predictor of behavioural intention was previously shown to be reliable to predict behavioural outcomes (Dinev et al., 2008). The research of Thong et al. (2006) has found a strong and reliable relationship between the individual’s attitude and their behaviour. However, behavioural adoption is not simply a comparison between the sum of outcomes regarding risks and benefits, as researchers such as Wang, Duong & Chen (2016) have shown that benefits have a stronger effect on the willingness to adopt.

Similarities and differences between adoption and continuance of use

(8)

of variance in motivations in the different stages of a technology usage lifecycle (Bhattacherjee, 2001 and Bhattacherjee and Premkumar, 2004; Davis, Bagozzi, & Warshaw, 1989; Venkatesh and Davis, 2000 and Venkatesh et al., 2003). Absence of understanding of these differences in behaviour over time will lead to ineffective management of privacy design, reputation building and organisational actions (Liao, Palvia & Chen, 2009). The past has shown that the viability of a technological innovation very much depends on the continued usage of individuals, implicating the importance of retention of the current customer base (Karahanna, 1999; Bhattacherjee, 2001).

The difference between adoption and post-adoption is that during adoption, expectations of perceived risks and benefits result in a particular attitude, determining the adoption behaviour. However, after adoption users no longer rely on expectations of perceived benefits and risks resulting in an attitude. Users are better able to give interpretation to their perceived benefits and risks, as they have experienced this during the use of the fitness tracker (Liao et al., 2009). This shows that the perception of risks and benefits during the use can significantly differ from the expectations of risks and benefits before adoption, giving potentially different outcomes between adoption and continuance of use. This is illustrated by Shen (2019) who found that during the adoption of technology, the focus of individuals is on the benefits of adoption, while during the use, the users becomes more aware of the risks and downsides of use (Li et al., 2016). Research conducted on the former does not necessarily give the same outcomes for the latter (Brown, Venkatesh & Goyal, 2014). More studies have to be conducted on continuance behaviour separately from adoption behaviour to have a better understanding of the differences and similarities. Due to this lack of separation and not treating continuance of use as a distinctive subject, there is the consequence that many findings in the literature about technology adoption are generalised to the later stages of the usage-lifecycle without any supported evidence (Thong et al., 2006; Liao et al., 2009).

The privacy calculus theory

(9)

information. Despite many individuals making an actual trade-off between the positives and negatives, this also often happens unconsciously (Li, 2014). Especially when comparing the users to adopters, the trade-off for users is more unconscious because the same trade-off is made every time the tracker is used, resulting in a less extensive overthinking of risks and benefits. Adopters will be more aware of this trade-off compared to users, as the individual is likely unfamiliar with the technology and leaks the experience to make a quick evaluation (Aarts, Verplanken & van Knippenberg, 1998).

The theory of the privacy calculus has been applied broadly in the research fields of information systems, information sharing & e-commerce, because these fields are almost inevitably involved in privacy issues (Dinev, Bellotto, Hart, Russo, Serra & Colautti, 2020). Researchers studying the willingness to share information often do not distinguish health-related information from other forms of information, despite the perceived privacy risk differs. The willingness to share information depends on the personality of the information that is being requested for disclosure (Wolf et al, 2015). Therefore, disclosure of health information will be separated from other forms of information in different contexts (Xu et al., 2009). The next sections will further discuss the constructs of the privacy calculus theory.

Perceived privacy risks

(10)

Perceived benefits

The aforementioned advancements in technology do not only lead to negative outcomes, as the increased capabilities lead to higher levels of perceived benefits. The benefits of disclosing personal information with fitness trackers come from better access to care, convenience, patient empowerment and a more effective and efficient healthcare system (Shen, 2019). As the technologies are advancing, the capabilities of fitness trackers are no longer limited to the number of steps a day or number of calories burned .The increasing capabilities make it possible to get feedback on sleep cycles, heartbeat or even the oxygen levels of the human body (Wolf et al., 2015). The increased data collection has the benefit that more personalised feedback can be provided to the user, resulting in higher benefits (Gupta, Patel & Greenes, 2016). Studies such as Li (2014) and Abdelhamid et al. (2017) have focused on the risks-side when attempting to find the effects of perceived privacy risk on individual’s behavioural intention. However, the actual decision often could not be explained with only knowledge about the perceived risks, as it ignores that behaviour intentions are also the consequence of perceived benefits of continuing. The experienced benefits were found to have a significant effect on the willingness, as focusing on the benefits could make individuals almost ignore the fact that they were also experiencing risks (Li et al., 2016; Wang et al., 2016). This displays the need to include benefits and to approach the behavioural intention as a trade-off between these two variables. When perceived benefits are low or entirely absent, the individual is more likely to discontinue the sharing of personal information because the perceived privacy risks prevail (Kehr, Kowatsch & Wentzel, 2015). In theory, the trade-off happens rational and should therefore lead to abandoning of use because the balancing provides a clear negative outcome, but the next section will discuss this rationality in more detail.

Rationality of the trade-off decision

(11)

understanding to make a good prediction of the risks and benefits shows irrational behaviour. This could result in higher levels of continuance than would be expected based on the theory (Shen, 2019). In such situations, this could have a weakening effect on the negative association between perceived privacy risks and continuance intention. However, overall users are expected to have at least a basic understanding of the perceived risks of personal information disclosure, as the use of technology for information exchange is becoming more and more intertwined in our lives (Kim et al., 2019). The assumption of the privacy calculus theory, individuals making rational choices, combined with newer theories showing the sometimes irrationality of behaviour is not expected to have significant consequences for this research, because the effect of irrationality seems stronger for actual behaviour, and not for intentions of behaviour (Li et al., 2016; Shen, 2019).

(12)

there is no information to be shared anymore. Discontinuance of use always remains an option (Park, 2015). So, if perceived risk is surpassing benefits, the logical consequence would be to abandon the device, and not to share personal information less often. Therefore, this CPM theory provides a critical view on the privacy calculus, which tends to focus on the disclosure side of the decision. According to the CPM, some users would intend to completely discontinue the use if the balance between advantages and disadvantages is negative, similar to the decision to not-adopt in the previous stage. This provides an interesting view to this research, as it could be questioned if individuals who have spent money on a fitness tracker are not experiencing some feelings of being ‘locked-in’, which gives a user the moral feeling to continue (Buchwald, Letner, Urbach & von Entreß-Fürsteneck, 2018).

Continuance of use intention

As this research extends the privacy calculus model, the focus of outcomes will not be on adoption but on continued use, where the individual has already adopted the fitness tracker. The continuance of use of fitness trackers is, unlike adoption intention, not measured for a single moment in time but requires information about the intention to continue the use over a longer period of time (Huang, 2013). The continuance intention becomes relevant directly after the individual has made the decision to adopt. To measure the effects of the privacy calculus for current users, information will be required about their continuance intention (Thong et al., 2006). This also implies that the theory to describe the adoption intention, the well-known Technology Acceptance Model (TAM) will not be used. The TAM focuses only on initial acceptance of a technology, assuming that a decision is made at a single moment in time (adoption/no adoption) (Huang, 2013). As mentioned in earlier sections, attitude is the determinant of adoption intention, which is also where the focus of the TAM lies. Attitude becomes a less important determinant when measuring the continuance intention. This shows why copying research on adoption cannot be justified to describe post-adoption behaviour (Liao et al., 2009). Continuance intention can be explained according to the Expectancy-Confirmation Model (ECM), which uses a longer time-horizon to describe the development of perceived risks and benefits. The ECM will be further discussed in the next paragraph.

(13)

et al., 2016). Simply said, the benefits of the fitness tracker are somewhat overestimated in the adoption stage, while risks are underestimated. The better balance of the trade-off is the consequence of increased awareness for negative outcomes experienced during the use of the fitness tracker (Kim et al., 2019). The initial focus on benefits is in line with what is written about the TAM for adoption, which indeed focuses on the advantages of adoption, not on the related disadvantages. Second, there is another important difference between the adoption and continuance of use. The continuance intention is influenced by the individual’s expectations about the technology, the pre-purchase expectations, and the assessment of satisfaction after the technology has been adopted (Lee & Kwon, 2011). Continuance outcome is influenced by the discrepancy between these initial expectations and the confirmation of these expectations. During the use of the technology the individual gets more experience with the product leading to the assessment of benefits (Thong et al., 2006). Comparing the pre-adoption expectations with the after-adoption experiences leads either to confirmation or disconfirmation of expectations. The theory explains that continuance is dependent on outcomes of satisfaction for a long-term perspective, influenced by the initial expectations made during the adoption process (Akter et al., 2013). Disconfirmation can have two reasons and consequences, as a technology can be rated below expectations, giving a feeling of dissatisfaction and making discontinuance more likely. But also, the product could be rated above expectations, providing a feeling of satisfaction leading to a higher intention to continue (Bhattacherjee, 2001). Users perceiving a lot of benefits during the use are, compared to users perceiving low benefits, more likely to be positively disconfirmed by the outcomes based on their expectations (Brown, 2014). Therefore, according to the ECM theory, higher perceived benefits will associate with stronger intentions to continue the use of the tracker. This process of expectation and confirmation is shown in figure 1.

(14)

The theory mentions three factors determining continuance: the level of satisfaction, the extent of confirmation of expectations and post-adoption expectations, in the form of perceived usefulness (Bhattacherjee, 2001b; Thong et al., 2006). Notice that this is not only about the confirmation or disconfirmation of expectations, as absolute outcomes of benefits independent of expectations is most important for the continuance decision. However, it shows that besides the amount of the benefits, other factors play a role in this stage. Focusing only on expectation and confirmation could imply that those users having low expectations about benefits and have these expectations confirmed would show higher continuance intention compared to users with high expectations about benefits being slightly disconfirmed as expectations were too high. However, it can provide reasoning for users perceiving low levels of benefits but showing high continuance intention, as they simply had low expectations of benefits before adoption (Mckinney, Yoon & Zahedi, 2002).

Hypotheses development

Based on information about the constructs of the privacy calculus mode, both variables will have an opposing effect on the intention to continue the use of a fitness tracker. The first variable of the privacy calculus model is perceived privacy risks. More perceived privacy risks will give users a stronger feeling that their personal information might be disclosed with those who are not authorised to have access to this information according to the theory about the privacy calculus (Li et al., 2016). In such situations, the users predict a certain loss or harm of their privacy when they continue to use the technology (Guo, Li & Wu, 2020). The stronger the perception of privacy being at risk, the higher the expected outcomes of privacy violation/abuse will be. So, perceived privacy risk is negatively associated with continuance intention.

(15)

Hypothesis 1: Perceived privacy risk is negatively associated with the intention to continue

the use of a fitness tracker.

The second variable of the privacy calculus model, perceived benefits, are the advantages experienced as a consequence of the use of a fitness tracker. Perceived benefits incentivise the user to continue the use, as the use brings the individual benefits. Such benefits are usually the reason for adoption. As we know by now, based on the theoretical background section, benefits are the building block of continuance intention (Bhattacherjee, 2001). Developments in the technology of fitness tracker have created opportunities to receive increasingly personalised data, which leads to more perceived benefits as the feedback better fits the needs of the user (Sun et al., 2015; Wang et al., 2016. This is because the level of personalisation is found to increase the experienced benefits of users. Discontinuance of use would end the benefits of fitness tracker use, as the benefits of use will no longer be experienced by the user due to discontinuance. As Wang et al. (2016) and Kim et al. (2019) described in their research, individuals tend to focus on the benefits of the privacy calculus.

Based on current knowledge, it is expected that there is a positive relationship between the perceived benefits and continuance intention for users of a fitness tracker. Because discontinuance of the fitness tracker will end the perceived benefits of use, those users perceiving higher levels of benefits are more incentivised to continue. Individuals perceiving low benefits are less motivated to continue with the disclosure of their personal information. As the perceived gains from continuance are low, users will feel less the need to continue because the consequence of discontinuance of a fitness tracker will have low effect on perceived benefits. The users perceiving high benefits will be highly motivated to continue, as the tracker has a large contribution to the daily life of the user (Li et al., 2016; Wang et al., 2016).

Hypothesis 2: Perceived benefits are positively associated with the intention to continue the

use of a fitness tracker.

The moderating effect of trust

(16)

purchase in an online environment (Pavlou & Gefen, 2004). However, the role of trust becomes especially interesting in an online environment where personal health information is shared (Bansal et al., 2010; Wang et al., 2016). Perceived privacy risks and benefits are based on the evaluation of fitness trackers in general trackers, not to the specific type of fitness tracker that is used by the individual (Dinev & Hart, 2006). Trust however relates to the level of trust in the fitness tracker vendor of the user. Trust reflects the willingness to be vulnerable based on the expectations about the future behaviour of another party (Gefen et al., 2003; Zhang et al., 2008). So, users assess the probability that what is perceived to be the outcome will actually be the outcome for their specific fitness tracker type (Akter et al., 2013). This means that despite the user might perceive high privacy risks related to the disclosure of information, he or she has the feeling that the vendor can be trusted and will act upon his/her interests (Kim et al., 2008). As a consequence, increasing levels of trust positively moderate the negative effect of perceived privacy risks on continuance intention. For perceived benefits and continuance intention, trust has a similar role. The perceived benefits follow from level of personalised data collection, which means that more collection of personalised data by the tracker brings the user higher level of benefits, as more specific feedback can be provided. However, this does not have any implication for the reliability and accuracy of the measurements (Dinev & Hart, 2006). It might be questioned by the user whether the information provided by the tracker can be trusted to be precise and correct, or whether the information is expected to be more like a rough guess (Shen, 2019). For the latter, the perceived benefits will lead to lower outcomes of continuance intention compared to the former, as the user trusts the data provided by the fitness tracker to be correct (Xu et al., 2011). Therefore, increasing level of trust strengthen the positive association between perceived benefits and continuance intention.

(17)

context, it is often closely related with the perception of the carefulness of the vendor (Blakeney, 1986; Gabarro, 1987). Especially reliability and safety are expected to be the reason for a moderating effect and competence is more important in the pre-purchase phase, as the evaluation of fitness tracker vendors is about the prediction of their behaviour. Reliability and safety are important in later stages, during the use of the tracker when users have more experience with the product and the vendor (Gefen et al., 2003; Li et al., 2014).

Based on the theoretical background, the hypotheses for the moderating effect of trust are developed. High trust has a positive moderating effect on the negative relationship between perceived privacy risks and continuance intention. This is because the negative effect of perceived privacy risks on continuance intention is tempered because users have the belief that despite the potential risks for privacy because of personal information disclosure, the particular vendor can be trusted not to violate privacy and to act confidential and secure with the data (Kehr et al., 2006). So, this implies that although there might be a perceived risk for privacy, the users do not expect their particular vendor to engage in behaviour that results in privacy violation, as the vendor can be trusted. Low trust in the vendor will give the user the feeling that the perceived privacy risks will indeed lead to privacy violation by the vendor (Dinev et al., 2006). Therefore, the individual will lack the confidence that the vendor will act upon the interest of the individual. Therefore, low levels of trust are expected to negatively moderate the relationship between perceived privacy risks and continuance intention (Akter et al., 2013).

Hypothesis 3: The level of trust positively moderates the relationship between perceived

privacy risks and continuance intention. This negative relationship is less pronounced when the level of trust is higher rather than lower.

(18)

the effect of perceived benefits and continuance intention. Low trust in the vendor will have the consequence that provided feedback (delivering the perceived benefits) by the fitness tracker is expected to be unreliable and incorrect for the management of health (Bhattacherjee, 2002; Gefen, 2002). This means that despite a potential high level of personalized data collection, leading to high perceived benefits, the individual lacks trust to use this data. This has the result that the positive relationship between perceived benefits and continuance intention is weakened by the fact that the user has the feeling that the provided information cannot be trusted (Akter et al., 2013). All hypotheses are shown below in the conceptual model in figure 2.

Hypothesis 4: The level of trust moderates the relationship between perceived benefits and

continuance intention. This positive relationship is pronounced when the level of trust is higher rather than lower.

(19)

METHODS

Research design

The hypotheses were tested among Dutch-speaking frequent users of fitness trackers. A quantitative survey was conducted to find evidence for the privacy calculus theory among users of fitness trackers. The survey was shared in two types of Facebook groups, with a minimum number of 250 members to decrease the chances that participants are closely related to each other and to ensure diversity among respondents (Ugander, Backstrom, Marlow & Kleinberg, 2012). It was distributed in neighbourhood groups and groups created to exchange information related to physical well-being and exercising. The neighbourhood groups were used to make sure there is enough diversity among respondents, as the selected neighbourhoods are expected to give a good representation of fitness tracker users. The second type of group was used to increase the chances of potential respondents having a fitness tracker, due to their affinity with physical activity or a healthy lifestyle. Due to this affinity, they are expected to have a good understanding of not only the advantages, but also the disadvantages of fitness trackers as they are assumed to be engaged with their device. The survey was available for respondents between the 15th of May until the 28th of May in 2020 and it was designed and collected using Qualtrics

(20)

with continuance among the devices. The survey also included a control question in the middle of the survey, to see whether the respondents paid attention when filling in the survey. To ensure that the different items can be combined into a single variable, the Cronbach’s alpha will be tested to see whether the items are reliable and allowed to be combined into one variable. All hypotheses are tested using SPSS 26.0.

Data collection

Out of the 304 respondents, 19 where filtered because they did not pass the control question in the survey. Of the remaining respondents, 37 were filtered because they answered no to the question whether they are a frequent fitness tracker user. Ultimately, the data of 248 respondents could be used for data collection. The questionnaire is presented in the appendix 2.

Independent variables

To measure the variables of the privacy calculus, perceived privacy risk and perceived benefit the questionnaire of Li et al. (2016) is used. This measurement instrument was used in multiple studies and was extensively tested in multiple research settings for convergent and discriminant validity to ensure construct validity. Questions about the perceived privacy risk are related to the disclosure of personal information. These questions are about the perceived risks of fitness trackers in general. Questions about the perceived benefits are related to the benefits the tracker gives them to have better understanding of their health and the ability to increase their current health situation, compared to a situation where the user did not possess a fitness tracker. The variables measure to what extent the provided information and feedback of the fitness tracker give a feeling of perceived benefits.

Moderator

(21)

continuously shared, some users will have a good understanding about reputation, quality of the security systems and role of ethics for their particular vendor.

Dependent variable

Continuance intention is measured on 5-point Likert scale based on the survey of Akter et al. (2013), who conducted research on the effects of quality and trust in information systems in healthcare and tested the effect on continuance intention of these systems for individuals. The goal is to test future use intention, which differs from the actual behaviour of individuals as the actual usage is not measured in this study. Continuance intention is measured on a two-item scale. Although more is seen as better, the lack of multidimensionality (either and individual continues or not) allows it to be captured with these questions (Eisinga, Grotenhuis & Pelzer, 2013; Diamantopoulos, Sarstedt, Fuchs, Wilczynski & Kaiser, 2012). As the capabilities of fitness trackers quickly advance, a question is included related to the continuance with the current device of the user instead of an alternative. If the user would prefer a new tracker or different brand, it could show a lack of intention to continue with the current device, while their actual intention is to continue using a fitness tracker. This could result in lower levels of continuance intention.

Control variables

Control variables included in this research are age, gender, level of education, intensity of use, the type of device and the type of Facebook groups (Li et al., 2016; Wang et al., 2016).

Age is measured as a ratio and is therefore collected in absolute numbers.

As mentioned, in the past the concerns related to the disclosure of personal information was higher than it is today, as the rise of technological age has decreased general perceptions of risk. This could be the result of lower concerns among younger aged people, lowering the overall perceived risk. Therefore, age is included as a control variable in order to see whether older adults show more perceived privacy risks than younger adults (Wang, Duong & Chen, 2016).

A binary measurement is used to test gender. Research conducted on the differences between men and women related to the perceived risks of online personal information disclosure have shown that in some studies women show more concerns than men. This difference in perception of risk can influence the intention to continue the use if a significant relation is found between the independent and dependent variable (Wang et al., 2016).

(22)

downsides of information disclosure, perceived benefits or trust in vendors of fitness trackers can substantially differ because of these educational differences (Wang et al., 2016).

Lastly, the intensity of use is measured based on a nominal level, in which individuals’ use of their fitness tracker ranges from at least once a week to everyday use. Different levels of intensity of use can correspond with different perceived benefits, as those who are more engaged with their device, using it every day, might respond to perceive more benefits from the use compared with those who use it every now and then, and therefore use their device less often (Attig et al., 2018; Guo, Li & Wu, 2020).

The development of fitness trackers led to the existence of many different devices related to size, colour but also the functionality of the tracker. This means that many trackers from different brands have different capabilities, which means that the level of personalized data collection could differ among respondents. Therefore, the type of device is collected to measure if there are significant differences between the perceived privacy risk or perceived benefits among users of different devices. The different devices are included as dummy variables in the analysis (Gupta et al., 2016.

As there are two different types of Facebook groups used for respondents, the neighbourhood groups and groups for people who like to be active or have a healthy lifestyle, dummy variables for both groups as they exist for different reasons. This is to measure whether there is a significant difference between these two groups. Those being more engaged with physical activity could perceive more benefits from the use or could ignore the perceived privacy risks more than ‘ordinary’ users (Ugander et al., 2012).

RESULTS

The results of the statistical analyses will be presented in this section. This section starts with the descriptive statistics, followed by the results of the linear regression tests for the hypotheses and lastly the test on the moderating effect of trust.

Descriptive statistics & correlations

(23)

was higher professional education. As expected, most individuals wear their fitness tracker every day, which fits the purpose of these devices, as they are also meant to collect data on for example sleep cycles. Those who do not make daily use of it most likely use it for their workouts, although not explicitly asked, they answer to use it no more than twice a week.

As there is a large variety of brands in the fitness tracker industry, respondents were asked to fill in their type of device to include dummy variables for the different types of trackers, in order to make sure there are no significant differences in outcomes between different trackers. This difference could be a consequence of the different designs and functions of the trackers, leading to different perceptions of risk or benefits. Dummies were also included to control for the effect of the two different type of Facebook groups. The outcomes of the tests show that there are no significant differences between the so-called neighbourhood groups and the physical active/health groups.

This was followed by a multicollinearity test to see whether the independent and control variables were not correlating, which gave satisfying VIF-scores between 1 and 1.1. This is followed by the check of the control variables. The control variables were uncorrelated with the dependent variable continuance intention, except the intensity of use. Intensity of use showed a significant correlation of 0.155 with continuance intention, at a 95% confidence interval. This implies that those using their tracking more are also showing higher intention to continue that use. However, the intensity of use was uncorrelated with the independent variables.

(24)
(25)

Hypotheses testing

Table 2 displays the regression results that are used for hypotheses testing. Model 1 shows the results for the control variables. In model 2, the independent variables perceived privacy risk and perceived benefits are added. For model 3, the moderation results of the interaction between perceived privacy risk and trust are included, followed by the interaction effect between perceived benefits and trust.

Hypothesis 1 tests the relationship between the perceived privacy risk and continuance of use. Based on the theory, a negative relationship was predicted, implicating that higher perceived privacy risks are negatively related to the intention of individuals to continue with the use of their fitness tracker. The results in table 3 show that there is indeed a significant negative relationship between the independent and dependent variable (ß =-.26, p < .01). Therefore, there is support for hypothesis 1.

Hypothesis 2 tests the relationship between perceived benefits lead and the intention to continue the use of fitness trackers. Expected is that more benefits have a positively relationship with continuance of use. The results show that the positive relationship is found between these variables (ß =.40, p < .01). The effect of hypothesis 2, the perceived benefits, has a much stronger effect on continuance than hypothesis 1, perceived privacy risks. This shows that respondents put more weight on the benefits side of the trade-off, instead of the risks. Based on these results, support for hypothesis 2 is found.

Hypothesis 3 tests the moderating effect of trust on the negative relationship between perceived privacy risk and continuance intention. It was expected that increasing levels of trust positively moderate the direct relationship between perceived privacy risk and continuance intention. The outcome is not significant (ß = -.38; 95% confidence interval = -.19 to .02). This shows that the level of trust does not significant moderate the relationship between perceived privacy risk and continuance intention.

(26)

Table 2. Regression results with interaction effect

Variable Model 1 Model 2 Model 3

Control variables Age Gender Education Intensity of use -.13 (.00) -.08 (.10) .02 (.04) .18** (.04) -.08 (.00) -.06 (.09) .05 (.04) .16** (.04) -.08 (.03) -.07 (.09) .05 (.04) .16** (.04) Fitness tracker type

Garmin Huawei AppleWatch Samsung Xiaomi Others Facebook group Independent variables -.05 (.10) -.08 .24 .04 (.13) -.09 (.13) .00 (.30) -.12 (.15) .08 (.08) -.04 (.09) -.04 (.22) -.00 (.11) -.08 (.11) .06 (.27) -.10 (.13) .08 (.07) -.03 (.09) -.04 (.22) .00 (.11) -.08 (.11) .05 (.27) -.09 (.13) .08 (.07)

Perceived privacy risks

(27)

(.06) (.31)

Interaction effect

Perceived privacy risk * Trust

Perceives benefits * Trust

Constant 4.22** (-.01) 3.10** (.46) -.38 (.05) .27 (.04) 3.16** (1.39) R2 Adjusted R2 .064 .025 .272 .235 .280 .236

Notes: N =248; * p < .05; ** p < .01 with 95% confidence interval

DISCUSSION

Theoretical implications

(28)

vendor of the users creates a feeling of safety and confidence, resulting in the idea that despite the perceived risks of using fitness tracker, the particular vendor behaves in the interest of the user. Trust in the vendor also gives a feeling that all the information provided by the tracker, resulting in the perceived benefits, can be trusted to be correct and precise, providing valuable feedback to the user (Kim et al., 2008; Shen., 2019).

After the survey was conducted among frequent users of fitness trackers, outcomes of analyses have shown that the effects of perceived privacy risks and benefits are significant and outcomes comparable to research of Li et al. (2016) on adoption of fitness trackers (Xu, Luo & Carroll, 2011; Dinev & Hart, 2006; Li 2014). The perceived privacy risks have a significant negative effect on the intention to continue the use of a fitness tracker, although the strength of the effect is weak. The other variable, perceived benefits, is found to have a significant positive relationship with continuance intention. The effect of benefits is stronger than the effect of risks, although it has still a weak to moderate effect. The hypotheses testing the moderating effect of trust on the variables of the privacy calculus were not significant, and therefore there is no evidence that the relationships are moderated by trust.

Some interesting findings will now be discussed in more detail. First of all, benefits of technology usually have a stronger effect on the willingness to adopt compared to the risks, as individuals prefer to focus on the benefits (Li et al., 2016). Outcomes of analysis have shown that the effect of benefits is almost twice the effect of risks on the continuance intention. This stronger effect of benefits in the adoption stage was because individuals look for reasons to adopt, instead of reason to not adopt. Therefore, focus of individuals is on the perceived benefits of adoption. This thesis has shown the same effects for continuance of use, however, for another reason. A possible explanation for this effect is given by Abdelhamid, Gaia & Sanders (2017), who argue that, historically, healthcare professionals have focused on the education of patients on the benefits of sharing personal health information, ignoring potential related risks. However, as the focus of this research was in a healthcare-related setting, it is only relevant for research conducted in similar contexts and is no explanation for similar results in other contexts (Shen, 2019).

(29)

those perceiving a negative outcome of the calculus should entirely withdraw, not use their device less often. Reasons for the absence of these outcomes might be found in the benefits of disclosure, which incentivises to continue, despite the perceived risk of continuing. However, the strength of the relationships of the privacy calculus and continuance intention can be logically explanation as well, as the effect of perceived benefits was stronger compared to perceived privacy risks. Therefore, this thesis has contributed to the field of HIT by providing an extension of the privacy calculus model. Findings are now not limited to the willingness to adopt but are also shown for the individuals who have adopted and are now in the stage of continuance of use.

Literature describes rising popularity of fitness trackers, due to the effectiveness of relatively small devices with a large diversity of options and possibilities (Xu et al., 2011). This provides reasoning for the overall high continuance intention, as individuals seem to perceive the disadvantages to be marginal compared to the advantages. Another explanation for the high continuance intention can be found in the ECM theory of Bhattacherjee (2001). The theory states that next to the absolute outcomes of the privacy calculus, the expectations before adoption and confirmation or disconfirmation of these expectations after adoption play a role. If the ultimate perception of satisfaction is higher than expected upfront, users are positively surprised by the performance of the fitness tracker, increasing the feeling of satisfaction. Additionally, the overall high continuance intention could explain the absence of a significant moderating effect of trust, because users have the intention to continue no matter the level of trust they have in their vendor.

Managerial implications

(30)

intention among users. Despite this could increase the perceived privacy risks among users, the effect of benefits on the continuance intention is stronger.

Limitations and further research

Despite the overall success of this research, this thesis has several limitations. First of all, no questions were asked to the respondents how long they are users of a fitness tracker. This could have the consequence that some users are recent adopters, which could have implications for the assessments of perceived privacy risks and perceived benefits during the use, due to limited experience. Second, the measurements were limited to the behavioural intentions of users, ignoring the actual behavioural outcomes. This is due to the feasibility and limited scope of this thesis. Outcomes of actual behaviour would increase the value of findings about this topic.

Further research should be conducted to test whether individuals not only show the intention to continue or discontinue, but whether the intentions lead to actual behaviour. Intention and actual action are not necessarily the base, but this thesis provides a useful base to further research the relationship between continuance intention and actual behaviour. As the research of Li et al. (2016) has shown, there is a significant difference between adoption intention and actual adoption.

Second, overall the outcomes of continuance intention were high, despite severe risk for privacy. The CPM theory argued that negative outcomes of the privacy calculus trade-off should lead to discontinuance of use, not to less intensive use. However, further research should be conducted to investigate if users will actually discontinue the use and why users that perceive severe risks often show high intention to continue with the use of the tracker. As the outcomes of this thesis seem contrary to the findings of the CPM, further research would create understanding whether users feel obligated to continue the use of their tracker. Focus should be on the reasons why users feel this obligation to continue. Users could potentially feel being locked-in or feel pressure to continue as money was spent on the device and discontinuance would lead to sunk costs. However, there might be underlying and less visible reasons of motivations to continue with the use of a fitness tracker.

(31)

CONCLUSION

This research gives answer to the research question: to what extent is the continuance of use of fitness trackers influenced by the perceived privacy risks and perceived benefits, and to what extent is this relationship moderated by trust? The goal was to extend the privacy calculus model for users of fitness trackers, as literature on privacy effects is, until now, often focused on the adoption of fitness trackers.

In these times, where we seem to share more than ever before goes hand in hand with the increasing risks of misuse of our personal information. The negative effect of perceived privacy risks on adoption intention was already shown, but now there is evidence for the negative effects of these perceived risks on the continuance intention for fitness trackers. This is potentially valuable information for vendors of fitness trackers because it was unknow how perceived privacy risks influence the intentions of users to continue, despite the fact that continuance intention has important implications for the viability of the market for fitness trackers. Therefore, this thesis contributes to the existing literature about the effects of the privacy calculus in a healthcare setting. Findings provide an extension of the existing privacy calculus on adoption intention of HIT.

(32)

REFERENCES

Aarts, H., Verplanken, B. & van Knippenberg, A. (1998). Predicting Behavior from Actions in the Past: Repeated Decision Making or a Matter of Habit? Journal of Applied Social

Psychology, 28 (15), 1355-1374.

Abdelhamid, M., Gaia, J., & Sanders, G. L. (2017). Putting the focus back on the patient: How privacy concerns affect personal health information sharing intentions. Journal of Medical Internet Research, 19(9), 1–12.

Angst, C. & Agarwal, R. (2009) Adoption of electronic health records in the presence of privacy concerns: the elaboration likelihood model and individual persuasion. MIS

Quarterly 33 (2) 339–370.

Attig, C., Karp, A. & Franke, T. (2018). User Diversity in the Motivation for Wearables Activity Tracking: A Predictor for Usage Intensity. Proceedings of the 20th Congress of

International Ergonomics Association. 431-440.

Bansal, G., Zahedi, F. M., & Gefen, D. (2010). The impact of personal dispositions on information sensitivity, privacy concern and trust in disclosing health information online. Decision Support Systems, 49(2), 138–150.

Ball, A., Ramim, M., & Levy, Y. (2015). Examining users’ personal information sharing awareness, habits, and practices in social networking sites and e-learning systems.

Online Journal of Applied Knowledge Management, 3(1), 180.

Bhattacherjee, A. (2001). Understanding information systems continuance: An Expectation- Confirmation Model. MIS Quarterly, 25, 351–370.

Bhattacherjee, A. & Premkumar, G. (2004). Understanding changes in belief and attitude toward information technology usage: a theoretical model and longitudinal test. MIS

Quarterly, 28 (2), 229–254.

Blakeney, R. (1986). A transactional view of the role of trust in organizational communication. Transactional Analysis Journal. 16(1) 95–98.

Brown, S.A., Venkatesh, V., & Goyal, S. (2014). Expectation confirmation in information systems research: A test of six competing models. MIS Quarterly, 38(3), 729–756.

(33)

Chipperfield, C., & Furnell, S. (2010). From security policy to practice: Sending the right messages. Computer Fraud & Security, 10(3).

Davis, F.D., Bagozzi, R.P., Warshaw, P.R., (1989). User acceptance of computer technology: a comparison of two theoretical models. Management Science, 35 (8), 982–1003.

Diamantopoulos, A., Sarstedt, M., Fuchs, C., Wilczynski, P., & Kaiser, S. (2012). Guidelines for choosing between multi-item and single-item scales for construct measurement: A predictive validity perspective. Journal of the Academy of Marketing Science, 40(3), 434–449. Dinev, T., Bellotto, M., Hart, P., Russo, V., Serra, I., & Colautti, C. (2006). Privacy calculus

model in e-commerce - A study of Italy and the United States. European Journal of

Information Systems, 15(4), 389–402.

Dinev T. & Hart P. (2006). An extended privacy calculus model for e-commerce transactions.

Information Systems Research, 17(1), 61-80.

Dinev, T., Hu, Q. & Yayla, A. (2008). Is there an on-line advertisers’ dilemma? A study of click fraud in the pay-per-click model. International Journal of Electronic Commerce, 13 (2), 29-60.

Dixon, P. (2005). Electronic health records and the National Health Information Network: patient choice, privacy, and security in digitalized environments.

Eisinga, R., Grotenhuis, M. & Pelzer, B. (2013). The reliability of a two-item scale: Pearson, Cronbach or Spearman-Brown. International Journal of Public Health.

Fereidooni, H., Frassetto, T., Miettinen, M., Sadeghi, A. & Conti, M. (2017). Fitness Trackers: Fit for Health but unfit for Security and Privacy. Institute of Electrical and

Electronic Engineers, 19-24.

Gabarro, J.J. (1987). The dynamics of taking charge.

Gefen, D. (2002). Reflections on the dimensions of trust and trustworthiness among online consumers. Advances in Information Systems, 33 (3).

Gefen, D., Karahanna, E. & Straub, D.W. (2003). Trust and TAM in online shopping: an integrated model. MIS quarterly, 27(1), 51-90.

(34)

Guo, J., Li, N., Wu, Y., & Cui, T. (2020). Examining help requests on social networking sites: Integrating privacy perception and privacy calculus perspectives. Electronic Commerce

Research and Applications, 39.

Hair, J. F., Black, W. C., Babin, B. J., & Anderson, R. E. (2010). Multivariate Data Analysis. London, UK: Pearson Education.

Huh, Y. & Kim, S. (2008). Do early adopters upgrade early? Role of post-adoption behaviour in the purchase of next-generation products. Journal of Business Research, 61(1), 40–60. Jarvenpaa SL, Tractinsky N, Vitale M. (2000). Consumer trust in an internet store.

Information Technology and Management 1(1–2), 45–71.

Jasperson, J., Carter, P. E., & Zmud, R. W. (2005). A comprehensive conceptualization of post-adoptive behaviors associated with information technology enabled work systems. MIS Quarterly, 29(3), 525–557.

Karahanna, E., Straub, D. W., & Chervany, N. L. (1999). Information technology adoption across time: A cross-sectional comparison of pre-adoption and post-adoption beliefs. MIS Quarterly: Management Information Systems, 23(2), 183–213.

Keith, M. J., Thompson, S. C., Hale, J., Lowry, P. B., & Greer, C. (2013). Information disclosure on mobile devices: Re-examining privacy calculus with actual user behaviour. International Journal of Human Computer Studies, 71(12), 1163–1173.

Kellermann AL, Jones SS. What it will take to achieve the as-yet-unfulfilled promises of health information technology. Health Affairs. 2013. 32 (1).

Kim, D., Park, K., Park, Y., & Ahn, J. H. (2019). Willingness to provide personal information: Perspective of privacy calculus in IoT services. Computers in Human

Behavior, 92, 273–281.

Krasnova, H., Veltri, N. F., & Gunther, O. (2012). Self-disclosure and privacy calculus on social networking sites: the role of culture: intercultural dynamics of privacy calculus.

Business & Information Systems Engineering, 4(3), 127–135.

Laufer, R.S. & Wolfe, M. (1997). Privacy as a concept and a social issue: a multidimensional developmental theory. Journal of Social Issues. 22–42.

Lee, Y., & Kwon, O. (2011). Intimacy, familiarity and continuance intention: An extended expectation-confirmation model in web-based services. Electronic Commerce Research and

(35)

Lee, M. K. O., & Turban, E. (2001). A Trust Model for Consumer Internet Shopping 2001. 6(1), 75–91.

Li, Y. (2014). The impact of disposition to privacy, website reputation and website familiarity on information privacy concerns. Decision Support Systems, 57(1), 343–354.

Li, H., Sarathy, R. & Xu, H. (2010). Understanding situational online information disclosure as a privacy calculus. Journal of Computer Information Systems, 62-71.

Li, H., Wu, J., Gao, Y., & Shi, Y. (2016). International Journal of Medical Informatics Examining individuals’ adoption of healthcare wearable devices: An empirical study from privacy calculus perspective. 88(555), 8–17.

Liao, C., Palvia, P. & Chen, J. (2009). Information Technology Adoption Behavior Life Cycle: Toward a Technology Continuance Theory (TCT). International Journal of Information

Management, 29 (4), 309-320.

Limayem, M., Cheung, C., & Chan, G. (2003). Explaining Information Systems Adoption and Post-Adoption: Toward an Integrative Model. Association for Information Systems. Marwick, A. E. & Boyd, D. (2011). I tweet honestly, I tweet passionately: Twitter users,

context collapse, and the imagined audience. NewMedia & Society, 13(1), 114–133. Mayer, R. C., Davis, J. H., and Schoorman, F. D. 1995. “An integrative model of

organizational trust,” The Academy of Management Review (20:3), pp. 709–734. McGregor, D.M. (1967). The professional manager. New York:McGraw-Hill

McKnight, D. H., Lankton, N., & Tripp, J. (2011). Social networking information disclosure and continuance intention: A disconnect. Proceedings of the 44th Hawaii international conference on system sciences, Kauai, HI (pp. 1–10).

McKnight, D.H., Choudhury, V. & Kacmar, C. (2002). Developing and validating trust measures for e-commerce: an integrative typology, Information Systems Research 13 (3) 334–359.

Oh H, Rizo C, Enkin M, Jadad A, Powell J, Pagliari C. (2005). What Is eHealth: A Systematic Review of Published Definitions. Journal of Medical Internet Research. 7(1).

Oliver, R. (1980). A cognitive model for the antecedents and consequences of satisfaction.

Journal of Marketing Research, 17, 460–469.

(36)

Patel, C. & Kushniruk, V. (2002). The patient clinical information system (PatCIS): technical solutions for and experience with giving patients access to their electronic medical records. International Journal of Medical Information. 68(3), 113–127.

Patterson, H.: Contextual expectations of privacy in self-generated health information flows.

TPRC 41(1), 1–48.

Pavlou, P.A, Gefen, D. (2005). Psychological contract violation in online marketplaces: antecedents, consequences, and moderating role. Information Systems Research 16(4), 372–434.

Petronio, S. (2010). Communication Privacy Management Theory: What Do We Know About Family Privacy Regulation? Journal of Family Theory & Review, 2(3), 175–196. Petronio, S. (1991). Communication boundary perspective: A model of managing the

disclosure of private information between marital couples. Communication Theory, 4, 311 – 332.

Petronio, S. (2010). Communication Privacy Management Theory: What Do We Know About Family Privacy Regulation? Journal of Family Theory & Review. 2(3), 175–196. Roger N. Blakeney (1986) A Transactional View of the Role of Trust in Organizational

Communication. Transactional Analysis Journal, 16:2, 95-98

Rust, R., Kannan, P. & Peng, N. (2002). The customer economics of internet privacy. Journal

of the Academy of Marketing Science, 30 (4), 455-464.

Smith, H. J., Dinev, T., & Xu, H. (2011). Information privacy research: An inter- disciplinary review. MIS Quarterly, 35(4), 989–1016.

Steinfeld, Nili (2015). “Trading with privacy: The price of personal information”. Online information review, 39(7), 923-938.

Shen, N. (2019). The e-health trust model: Understanding the patient privacy perspective in a digital health environment.

Shin, D. & Hwang, Y. (2017). Integrated acceptance and sustainability evaluation of Internet of Medical Things: A dual-level analysis. Journal of Internet Research, 27 (5), 1227-1254. Son, M., & Han, K. (2011). Beyond the technology adoption: Technology readiness effects on

(37)

Sun, Y., Wang, N., Shen, X. L., & Zhang, J. X. (2015). Location information disclosure in location-based social network services: Privacy calculus, benefit structure, and gender differences. Computers in Human Behaviour, 52, 278–292.

Tan, Y. H., & Thoen, W. (2000). INCAS: A legal expert system for contract terms in electronic commerce. Decision Support Systems. 29(4), 389–411.

Thong, J. Y. L., Hong, S. J., & Tam, K. Y. (2006). The effects of post-adoption beliefs on the expectation-confirmation model for information technology continuance. International

Journal of Human Computer Studies, 64(9), 799–810.

Venkatesh, V., & Davis, F. (2000). A Theoretical Extension of the Technology Acceptance Model: Four Longitudinal Field Studies. Management Science 46(2), 186-204.

Venkatesh, V., Morris, M., Davis G. & Davis F. (2003). User acceptance of information technology: toward a unified view. MIS Quarterly, 27 (3), 425-478.

Wang, T., Duong, T. D., & Chen, C. C. (2016). Intention to disclose personal information via mobile applications: A privacy calculus perspective. International Journal of Information

Management, 36(4), 531–542.

Wolf, C., Polonetsky, J. & Finch, K. (2015). A Practical Privacy Paradigm for Wearables.

Future of privacy platform.

Wolf, C. (2013). Wearable in the workplace. Harvard Business Review. 91, 23–25.

Xu, H., Teo, H., Tan, B. & Agarwal, R. (2009). The role of push–pull technology in privacy calculus: the case of location-based services, Journal of Management and Information Systems. 26, 135–173.

Yoon, C., & Rolland, E. (2015). Understanding continuance use in social networking services.

Journal of Computer Information Systems, 55(2), 1–8.

(38)

APPENDIX

Appendix 1: Measurement items of research variables.

Construct Items Description

Perceived Privacy Risk Perceived Benefits Trust Continuance Intention PPR 1 PPR 2 PPR 3 PB 1 PB 2 PB 3 T 1 T 2 T 3 CI 1 CI 2

I find it risky to disclose my personal health information to wearable devices vendors

There is high potential for loss associated with disclosing my personal health information to vendors providing wearable devices

There is too much uncertainty associated with giving my personal health information to vendors producing wearable devices

Using a wearable device improves my access to health information Using a wearable device improves my ability to manage my health Using a wearable device improves the quality of my healthcare I believe that the seller of my tracker is trustworthy

The seller gives me the expression that it keeps promises and commitments

I believe that the seller has my best interests in mind

(39)

Appendix 2: Questionnaire

General questions

1. What is your age?

……… years old

2. What is your gender?

[ ] Male [ ] Female

3. What is your highest level of completed education?

[ ] High school [ ] Mbo

[ ] Hbo [ ] University

4. I am currently using a wearable for fitness tracking

[ ] No

(40)

Perceived privacy risk

1. It would be risky to disclose my personal health information to wearable devices

vendors. [ ] strongly disagree [ ] disagree [ ] neither disagree/agree [ ] agree [ ] strongly agree

2. There would be high potential for loss associated with disclosing my personal health

information to vendors providing wearable devices. [ ] strongly disagree

[ ] disagree

[ ] neither disagree/agree [ ] agree

[ ] strongly agree

3. There would be too much uncertainty associated with giving my personal health

information to vendor providing wearable devices. [ ] strongly disagree

[ ] disagree

[ ] neither disagree/agree [ ] agree

(41)

Perceived benefits

1. Using a wearable device would improve my access to my health information.

[ ] strongly disagree [ ] disagree

[ ] neither disagree/agree [ ] agree

[ ] strongly agree

2. Using a wearable device would improve my ability to manage my health.

[ ] strongly disagree [ ] disagree

[ ] neither disagree/agree [ ] agree

[ ] strongly agree

3. Using a wearable device would improve the quality of my healthcare.

[ ] strongly disagree [ ] disagree

[ ] neither disagree/agree [ ] agree

(42)

Trust

1. I believe that the seller of my tracker is trustworthy.

[ ] strongly disagree [ ] disagree

[ ] neither disagree/agree [ ] agree

[ ] strongly agree

2. The seller gives me the expression that it keeps promises and commitments.

[ ] strongly disagree [ ] disagree

[ ] neither disagree/agree [ ] agree

[ ] strongly agree

3. I believe that the seller has my best interests in mind.

[ ] strongly disagree [ ] disagree

[ ] neither disagree/agree [ ] agree

(43)

Continued intention to use

1. I intend to continue using my wearable device to get health information.

[ ] strongly disagree [ ] disagree

[ ] neither disagree/agree [ ] agree

[ ] strongly agree

2. My intentions are to continue using wearables rather than using any alternative.

[ ] strongly disagree [ ] disagree

[ ] neither disagree/agree [ ] agree

Referenties

GERELATEERDE DOCUMENTEN

(a) Brian: All Entities (b) Brian: Wiki Entities (c) Brian: Non-Wiki Entities Figure 3: Disambiguation results over different top k frequent terms added from targeted tweets.. the

The ethicist, philosopher, or social scientist engaged in this type of Ethical Technology Assessment has the role of improving the conditions for deliberation in which

It states that there will be significant limitations on government efforts to create the desired numbers and types of skilled manpower, for interventionism of

As there is significant coupling between resonances 1 and 2 in the unpumped sample, to extract the bare resonance frequencies we used only data taken at pump powers &gt;10 µW.

Indicates that the post office has been closed.. ; Dul aan dat die padvervoerdiens

In addition, in the first part of the questionnaire, respondents were asked to provide the name of a specific retailer they had a personal omni-channel experience with (using both an

They also count with movement reminders (in the form of device vibration after certain periods of physical inactivity), personally tailored routines and even

privacy!seal,!the!way!of!informing!the!customers!about!the!privacy!policy!and!the!type!of!privacy!seal!(e.g.! institutional,! security! provider! seal,! privacy! and! data!